"Time-Series Data for Biogas Production and Lag-Based Prediction in a Swine Farm"

"This repository provides the dataset and structured data subsets used for the predictive modeling of biogas production in a swine farm bio-digester, as presented in the paper \"Biogas Prediction Enhancement of a Swine Farm Bio-Digester Using a Lag-Based Surrogate Machine Learning Model\". The data is specifically curated to facilitate the application and validation of machine learning algorithms in anaerobic digestion processes.A core contribution of this repository is the inclusion of distinct datasets structured for evaluating both a base model and a lag-based surrogate model. To support the sensitivity analysis discussed in the paper, the data is organized into folders containing varying numbers of registers. These subsets allow for the replication of the study's further analysis, demonstrating how the predictive performance (measured by the R^2 score) plateaus after reaching a specific threshold of data points.Researchers can utilize these datasets not only to reproduce the findings regarding the data size performance limit but also to benchmark new forecasting architectures and optimize data collection strategies in full-scale biogas plants."